AI · Studies · 2026-10-18 09:30
This open-source benchmark simulates agent behavior to compare search engine APIs across five verticals including news and legal, with blind ranking based on titles and snippets, without page fetching or reranking.
Auto-generated summary for reference only. RSS